Towertalk
[Top] [All Lists]

[TowerTalk] Antenna proximity--problems with rx overload

To: <towertalk@contesting.com>
Subject: [TowerTalk] Antenna proximity--problems with rx overload
From: W8JI@contesting.com (Tom Rauch)
Date: Sun, 7 Jan 2001 09:06:38 -0500
Hi Tom and all,

> To protect my 830's, I now have 'protection boxes'
> in series with the RX Antenna line (available through
> the transverter plug).  These boxes consist of a 
> series 2W carbon composition (non-inductive)
> resistor followed by a  pair of 4 high PIN diode
> stacks of opposite polarity.  Any low capacitance
> high speed switching diode should also work.

PIN diodes are actually slow diodes, not fast. They "hold" in-to or 
out-of conduction for longer periods than normal switching diodes. 
That's why a small dc bias "holds" them on or off for fast changing 
RF. The diodes just don't see the RF, because they are too slow. 
They are sensitive to dc bias, not RF voltage, when operated over 
their normal frequency range. 

You'd never want to use an actual PIN diode for an RF clamp 
unless you apply dc bias before you apply the RF. By the way, 
these diodes are called PIN (and engineers mostly hate the term 
"pin" because it is a P I N diode). The name comes from the 
intrinsically pure silicon wafer they are made from. That's why they 
are expensive, they come from a pure and thicker than normal 
silicon wafer. 

The best diode for a self-biased clamp are just regular fast 
switching diodes. I use 1N914's or fast switching diodes when I use 
them, and if power levels are high I use a small 12 volt light bulb 
between the back-to-back diodes and the antenna lead. 

Do NOT use PIN diodes, and do not use Schottky devices (they 
breakdown at too low a voltage)!

Also remember these diodes can be the source of IMD and TVI! If 
they are clamping, they are mixing and generating harmonics.

> hard to find these days but other non-inductive
> resistors are available.

The other resistors are actually better. Carbon has a tendency to 
age down in value. Being a semiconductor, carbon resistors can 
"short". If you use a metal-film resistor the failure mode is always 
open.

That's a trick in designing stuff. You look at the resistor in a critical 
application, and decide whether a short or open would be the better 
failure mode. If you want a resistor that is more likely to fail 
shorted, you use a heavy carbon comp. If it can never fail shorted 
without hurting something, you use a metal type.

I use a series lamp (you may have to parallel a few to get the 
resistance low enough) because it increases resistance with power 
and it acts like a fast fuse. If the diodes are really getting whacked 
with RF a series lamp reduces harmonics radiated back into the 
RX antenna more than a fixed value resistor would.

If you have trouble with BC harmonics and stuff, you might want to 
series connect a string of back-to-back parallel diodes and 
increase the clamping voltage.
73, Tom W8JI
w8ji@contesting.com

--
FAQ on WWW:               http://www.contesting.com/FAQ/towertalk
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com


<Prev in Thread] Current Thread [Next in Thread>